Computational tools for mechanistic discrimination in the reductive and metathesis coupling reactions mediated by titanium(IV) isopropoxide

Article abstract of DOI:10.1007/s12039-012-0333-2

A theoretical study has been carried out at the B3LYP/LANL2DZ level to compare the reactivity of phenyl isocyanate and phenyl isothiocyanate towards titanium(IV) alkoxides. Isocyanates are shown to favour both mono insertion and double insertion reactions. Double insertion in a head-to-tail fashion is shown to be more exothermic than double insertion in a head-to-head fashion. The head-to-head double insertion leads to the metathesis product, a carbodiimide, after the extrusion of carbon dioxide. In the case of phenyl isothiocyanate, calculations favour the formation of only mono insertion products. Formation of a double insertion product is highly unfavourable. Further, these studies indicate that the reverse reaction involving the metathesis of N,N ^′-diphenyl carbodiimide with carbon dioxide is likely to proceed more efficiently than the metathesis reaction with carbon disulphide. This is in excellent agreement with experimental results as metathesis with carbon disulphide fails to occur. In a second study, multilayer MM/QM calculations are carried out on intermediates generated from reduction of titanium(IV) alkoxides to investigate the effect of alkoxy bridging on the reactivity of multinuclear Ti species. Bimolecular coupling of imines initiated by Ti(III) species leads to a mixture of diastereomers and not diastereoselective coupling of the imine. However if the reaction is carried out by a trimeric biradical species, diastereoselective coupling of the imine is predicted. The presence of alkoxy bridges greatly favours the formation of the d,l (±) isomer, whereas the intermediate without alkoxy bridges favours the more stable meso isomer. As a bridged trimeric species, stabilized by bridging alkoxy groups, correctly explains the diastereoselective reaction, it is the most likely intermediate in the reaction.

Full text of DOI:10.1007/s12039-012-0333-2

J. Chem. Sci. Vol. 124, No. 6, November 2012, pp. 1343–1352. ꢀc Indian Academy of Sciences.
Computational tools for mechanistic discrimination in the reductive
and metathesis coupling reactions mediated by titanium(IV) isopropoxide
∗
AKSHAI KUMAR and ASHOKA G SAMUELSON
Department of Inorganic and Physical Chemistry, Indian Institute of Science, Bangalore 560 012, India
e-mail: ashoka@ipc.iisc.ernet.in
Abstract. A theoretical study has been carried out at the B3LYP/LANL2DZ level to compare the reactivity of
phenyl isocyanate and phenyl isothiocyanate towards titanium(IV) alkoxides. Isocyanates are shown to favour
both mono insertion and double insertion reactions. Double insertion in a head-to-tail fashion is shown to be
more exothermic than double insertion in a head-to-head fashion. The head-to-head double insertion leads to the
metathesis product, a carbodiimide, after the extrusion of carbon dioxide. In the case of phenyl isothiocyanate,
calculations favour the formation of only mono insertion products. Formation of a double insertion product is
ꢁ
highly unfavourable. Further, these studies indicate that the reverse reaction involving the metathesis of N,N -
diphenyl carbodiimide with carbon dioxide is likely to proceed more efficiently than the metathesis reaction
with carbon disulphide. This is in excellent agreement with experimental results as metathesis with carbon
disulphide fails to occur. In a second study, multilayer MM/QM calculations are carried out on intermediates
generated from reduction of titanium(IV) alkoxides to investigate the effect of alkoxy bridging on the reactivity
of multinuclear Ti species. Bimolecular coupling of imines initiated by Ti(III) species leads to a mixture of
diastereomers and not diastereoselective coupling of the imine. However if the reaction is carried out by a
trimeric biradical species, diastereoselective coupling of the imine is predicted. The presence of alkoxy bridges
greatly favours the formation of the d,l (± ) isomer, whereas the intermediate without alkoxy bridges favours
the more stable meso isomer. As a bridged trimeric species, stabilized by bridging alkoxy groups, correctly
explains the diastereoselective reaction, it is the most likely intermediate in the reaction.
Keywords. DFT; hybrid methods; QM/MM; ONIOM; phenyl isocyanate; phenyl isothiocyanate and
titanium(IV) alkoxide.
1
. Introduction
Earlier these methods were limited to only small
molecules and could not be used to address complex
systems such as proteins and biomolecules. How-
3
The discovery of new catalytic processes depends to
a great extent on serendipity. However, if one has to
design and develop efficient catalytic processes, one has
to understand the structure, property, reactivity and sta-
bility of the reactants, products and intermediates in
the reaction. While the structure and reactivity of sta-
ble reactants and complex catalysts are often accessi-
ble through modern analytical tools, such experimental
methods fail to provide reliable information about tran-
sient species involved in a reaction. Recent advances
in theoretical studies, and enormous improvement in
computational power, enable good estimations of the
stability and structure of the species involved in chemi-
ever hybrid theoretical approaches have been formu-
lated in recent times, where a large molecular system is
divided into several subsystems or layers and each layer
is treated at different levels of computational detail.4
In this approach, only the important part of the sys-
tem which is directly involved in the reaction (such
as bond making and bond breaking) is computation-
ally evaluated with rigor. The other parts of the system
which are not involved in bond making or breaking are
handled using a lower and less demanding computa-
tional method. For instance, in the now well established
multilayer QM/MM method, a complex molecule is
divided into two parts, a high layer and a low layer.
The high layer is computed in a quantum mechanical
fashion whereas the low layer is computationally evalu-
1
cal reactions. Significant inroads have been made into
the understanding and prediction of intermediates in
chemical transformations at a molecular level.
2
5
ated using molecular mechanics. These hybrid com-
putational methods have been successfully employed
in transition metal chemistry to study the influence of
∗
For correspondence
1343

1344
Akshai Kumar and Ashoka G Samuelson
ligand environments on the catalytic activity of metal diastereoselectivity in coupling imines mediated by
13d
complexes. ‘Real’ molecules can now be treated more low valent titanium (LVT) reagents. A multinuclear
accurately and one can obtain more insight from these species, presumably a trimer, was said to be involved and
hybrid approaches. This allows one to correctly predict was said to be responsible for the observed diastereo-
the influence of a ligand, tailor its environment around selectivity.
a metal centre, and generate more efficient catalysts.
Computational methods have thus paved the way for to elucidate the reactivity of phenyl isocyanate and
advancements in catalysis in a rational way.6
phenyl isothiocyanate towards mono and double inser-
In the present study, we extended our investigations
There have been numerous reports where com- tion into titanium(IV) alkoxides. The mono and double
putational methods have been efficiently used to insertion of phenyl isocyanate into the Ti–OR bond is
elucidate the reactivity patterns of group(IV) metal calculated to be a facile process. However in the case
7
complexes. Jemmis and co-workers have employed of phenyl isothiocyanate only mono insertion appears
molecular orbital studies using extended Huckel and to be favourable. The double insertion and the subse-
MNDO methods to explain the difference in reacti- quent metathesis reaction of phenyl isothiocyanate is
8
vity of group(IV) metallocenes towards bis acetylenes.
computed to be energetically unfavourable. Secondly,
Recently, the groups of Jemmis and Rosenthal have we carry out QM/MM calculations to compare the
reported DFT studies at either the BP86 or the stability of multinuclear Ti-alkoxy species which are
BS1:LANL2DZ levels to explain the reactivity of bridged and non-bridged. It is shown that the interme-
group(IV) metallocene alkyne complexes with carbodi- diate that best explains the observed diasteroselectivity
9
imides. The formation of a five-membered hetero- is a trimeric alkoxy bridged species. Non-bridged forms
metallacycloallenes in the reaction of metallocene and bridged dimeric species are predicted to result in
alkyne complexes with carbodiimides has been postu- loss of diasteroselectivity.
lated by DFT methods and confirmed by experimental
techniques. The research group of Philip Mountford has
extensively used DFT methods to explain the diver-
2
. Results and discussion
10
sity in reactions of titanium hydrazido and titanium
11
imido complexes towards insertion/cycloaddition of
unsaturated substrates such as nitriles, heterocumu-
lenes and olefins. Our group has carried out experimen-
tal and computational investigations on the insertion
2.1 Comparison of insertion and metathesis
reactivity of isocyanates and isothiocyanates
catalysed by titanium (IV) alkoxides
and metathesis reactivity of heterocumulenes and hete- Earlier reports indicate that the insertion of tita-
1
2
roalkenes mediated by group(IV) metal complexes.
nium(IV) alkoxides and zirconium(IV) alkoxides into
In these reports, the differences in reactivity of Ti and heterocumulenes are temperature dependent.12 Hete-
its congeners towards insertion and metathesis have rocumulenes such as isocyanates 2O undergo mono
been explained by DFT calculations carried out at insertion and head-to-tail double insertion at room tem-
the B3LYP/LANL2DZ level of theory. Several organic perature. The mono and double insertion products are
ꢁ
ꢁ
reactions mediated by Ti complexes have been com- isolated as 3 aO and 4 aO after hydrolysis (scheme 1).
putationally studied using QM/MM methods.1 More However, at higher temperatures (110 C) in addition to
recently, we employed multilayer ONIOM calculations mono insertion and head-to-tail double insertion pro-
using QM/MM methods to satisfactorily explain the ducts, head-to-head double insertion products are also
3
◦
Scheme 1. Reaction of 2O/2S with titanium(IV) alkoxides at room temperature
12
leading to insertion products after aqueous work up.

Computational tools.....reactions mediated by titanium(IV) isopropoxide
1345
Scheme 2. Insertion-cum-metathesis products isolated after hydrolysis in the reac-
◦
12
tion of 2O/2S with titanium(IV) alkoxides at 110 C.
observed. Such a double insertion of aryl isocyanate in DFT calculations carried out at the (B3LYP/
a head-to-head fashion results in the elimination of car- LANL2DZ) level of theory are helpful in explain-
bon dioxide leading to metathesis. Stoichiometric reac- ing why isothiocyanates fail to undergo head-to-tail
tions lead to isolation of metathesis-cum-insertion pro- and head-to-head double insertion. To reduce com-
n
ducts 6a (scheme 2). The metathesis reaction are shown putational load, Ti(O Bu) is modelled by Ti(OMe)
4
4
to be reversible; forming unsymmetrical carbodiimides where bulky n-butoxide groups are replaced by methox-
12b,12g
when carried out with different aryl isocyanates.
ide groups. The intermediates involved numbered and
It is also observed that steric factors slow down the given codes depending on the “OR” group present (a:
n
metathesis reaction which is similar to the observation R = Bu, b: R = Me) and heterocumulene (PhNCX)
with insertion reactions. However, contrary to the sub- employed (O: X = O, S: X = S) in the reaction. For
stituent effects in insertion reactions, where electron instance, the complex 4aOHH indicates structure 4
n
withdrawing groups on the isocyanate favour the reac- with O Bu as the alkoxy group and a structure where
1
2b,12g
tion,
donating groups.
electronic effects, the changes on moving from Ti to Zr with OMe as the alkoxy group and a structure where
metathesis reactions are favoured by electron PhNCO has undergone double insertion in a head-to-
12b,12e,12g,12h
Similar to the reversal of head fashion. Similarly, 4bSHT indicates structure 4
also show a reversal.1
2d,12e,12f,12h,
Zirconium alkoxides PhNCS has undergone double insertion in a head-to-tail
are shown to be better catalysts than titanium alkoxides fashion.
for the metathesis reaction.1
2f
Titanium(IV) methoxide reacts with the first
One expects similar reactions and reactivity patterns molecule of PhNCX (2O/2S) to generate the metal
with isothiocyanates 2S. Surprisingly, the reactivity pat- carbamate/thiocarbamate intermediate 3bO/3bS. This
1
2
tern observed with aryl isothiocyanates is different.
reaction is computed to be more exothermic for
ꢁ
Only the mono insertion product 3 aS is observed. The PhNCO compared to PhNCS (scheme 3). However,
double insertion product 4 aS is not formed. Even at the reaction of 2S with 1b is endothermic by only
ꢁ
elevated temperatures, the metathesis products 6a and 3 kcal/mol. This barrier may be not high enough to
6
are not formed. It is postulated that the sulphur prevent the formation of 3bS. Hence one would expect
on the thiocarbamate 3aS (scheme 3) is not nucleo- to observe mono insertion products of both 2O and
philic enough to attack the incoming isothiocyanate to 2S. This is in agreement with experimental results
form the six-membered intermediate 4aSHH or 4aSHT which show formation of both mono insertion products
12
(scheme 4).
3aO/3aS.
Scheme 3. Insertion of 2O/2S into Ti–OR bond resulting in formation of metal carbamate species 3aO/3aS/3bO/3bS.

1346
Akshai Kumar and Ashoka G Samuelson
Scheme 4. Second insertion of 2O/2S into the metal carbamate species 3aO/3aS/
3bO/3bS in a head-to-tail (path Y) or head-to-head (path Z) fashion.
Subsequent insertion of PhNCX (2O/2S) can is high enough to hinder the formation of 4bSHH. This
take place either in a head-to-tail fashion giving renders the second insertion of 2S into 3bS in a head-
bOHT/4bSHT (Path Y) or in a head-to-head fash- to-head mode unfavourable. This argument is consis-
ion to give the intermediate 4bOHH/4bSHH (Path Z) tent with experimental results where 2S does not lead to
scheme 4). Experimental results indicate that the dou- metathesis products 6a/6.
ble insertion proceeds only though Path Y at room tem- The extrusion of carbon dioxide from 4bOHH to
4
(
perature. Further, Path Y is observed only with phenyl give the carbodiimide inserted intermediate 5b is less
isocyanate 2O and not with phenyl isothiocyanate 2S. exothermic compared to the extrusion of carbon disul-
In a similar manner, at higher temperatures head-to- phide from 4bSHH to give 5b (scheme 5, figure 2).
head insertion occurs in greater amounts through Path Hence it is likely that it is the previous step, 3bS→
Z leading to metathesis. While 2O underwent this 4bSHH, that has a detrimental effect on the metathesis
metathesis reaction, 2S failed to exhibit such a head- of 2S.
to-head double insertion reaction. The reluctance of
One could assume that the reverse reaction involving
isothiocyanates to undergo double insertion either in a the metathesis of carbon dioxide/carbon disulphide with
head-to-tail fashion or in a head-to-head fashion can carabodiimide 6 to give the isocyanate 2O/2S via the
be understood if one looks at the energetics of this intermediate 4bOHH/4bSHH follows the same path.
reaction.
For PhNCO, path Y is an exothermic process,
In this case, the process 5b→ 4bSHH is energetically
whereas path Z is endothermic (scheme 4). This
explains why head-to-head insertion is not observed at
room temperature. For path Y it is seen that, the for-
mation of 4bOHT is an exothermic process. Hence
the head-to-tail double insertion occurs readily and the
ꢁ
intermediate 4aOHT leads to the isolation of 4 aO
after hydrolysis. On the other hand, the formation of
4
bSHT from 3bS is calculated to be highly endother-
mic. The intermediate 4bSHT is destabilized by about
2 kcal/mol compared to 1b + 2S (figure 1). Hence the
double insertion in a head-to-tail mode occurs only with
1
2O and not with 2S.
Path Z is endothermic for both 2O and 2S, however
the formation of 4bOHH is less endothermic. The reac-
tion of 3bS with 2S in a head-to-head fashion leading
to 4bSHH appears to be a highly unfavourable pro-
cess. The species 4bSHH is destabilized by 24 kcal/mol
Figure 1. Computed energy profile for the sequential
insertion of 2O/2S into Ti(OMe)4 through path Y done at the
compared to the alkoxide 1b. It is likely that this energy B3LYP/LANL2DZ level of theory.

Computational tools.....reactions mediated by titanium(IV) isopropoxide
1347
Scheme 5. Reversible extrusion of CO2/CS2 from
aOHH/4aSHH/4bOHH/4bSHH.
4
more costly compared to 5b→ 4bOHH (scheme 5,
figure 2). Hence, the metathesis of carbon dioxide is
also better than the metathesis of carbon disulphide. Not
surprisingly, this is confirmed by experimental results
where carbon disulphide fails to undergo a metathe-
sis reaction with carbodiimide. The steps involved
in the catalytic cycle are given in scheme 6. The
overall energy changes associated with this catalytic
reaction is depicted in figure 2. Thus, the computed
thermodynamic stabilities of intermediates provide
useful insights into the contrasting preferences of
PhNCO and PhNCS towards double insertion and sub-
sequent extrusion of carbon dioxide and carbon disul-
phide, respectively.
Scheme 6. Intermediates involved in the plausible
metathesis of 2O/2S mediated by titanium(IV) alkoxide.
species, such as 8 (shown in scheme 7). Species 7 can
form dimers such as 9 and 9B.
Reduction of the imine by these species leads sub-
sequently to an amine 12 or coupling of the imine can
13
produce diammines 13 (scheme 8). Although reduc-
tion and coupling of imines occurred with both reduc-
ing agents, the coupling of imines is diastereoselective
when phenylsilane is used as a reducing agent. This
suggests that the nature of the LVT species formed
with a Grignard and the intermediate generated with the
imine is different from what is formed with phenylsi-
lane.
The diastereoselective products obtained in the case
of reactions with phenylsilane, increased with the con-
centration of titanium(IV) isopropoxide used. This
clearly rules out the possibility of an unimolecular Ti
species as an intermediate in the coupling reaction. The
species that is likely to be involved is either a Ti(III)
species like 8 or a trimeric biradical species such as
2
.2 Comparison of the stability of bridged and
non-bridged intermediates generated by the reaction of
LVT titanium alkoxides and imines
When treated with reducing agents such as Grignard
reagents or phenylsilane, titanium(IV) alkoxides gene-
rate low-valent titanium species (LVT).1
3c,13d
The LVT
species can be either Ti(II) speicies, such as 7, or Ti(III)
9
which can reduce the imine. Mass spectral studies
showed the possible involvement of a trimeric biradical
species experimentally. But it is difficult to arrive at the
actual structure of the species involved in the reaction
from mass spectral data. To understand the observed
diastereoselectivity in the C–C coupling reaction, the
possible intermediates involved in the C–C bond for-
mation are modelled computationally. It is possible
that the isopropoxide ligands have some influence on
the observed selectivity, though these groups are not
directly involved in the reaction. Hence instead of sim-
Figure 2. Computed energy profile for the sequential
insertion of 2O/2S into Ti(OMe)4 through path Z estimated
at the B3LYP/LANL2DZ level of theory.
i
plifying the system by replacing the Pr groups on Ti

1348
Akshai Kumar and Ashoka G Samuelson
Scheme 7. Possible LVT species generated by the reaction of titanium(IV) isopropoxide with reducing agents.
Scheme 8. Reductive reactions of imine mediated by LVT generated from Grignard and phenylsilane.
Scheme 9. TiIV Intermediate 14 formed in the reaction of low valent titanium species 8 with imine 10.
Scheme 10. Formation of trimeric Ti species 15/15B.

Computational tools.....reactions mediated by titanium(IV) isopropoxide
1349
Scheme 12. The most favourable conformations formed of
7 (± ) and 17 meso.
1
Scheme 11. The most favourable conformers of 16 (± ) and
6 meso.
1
1
6B-(± ) gauche/16B-meso gauche are also optimized
and their relative energy estimated. All conformers,
by OMe groups, we use a QM/MM approach to evalu- gauche, anti and eclipsed forms of the two isomers
ate the relative stabilities of the two diastereomers. are examined. The relative energies of all the opti-
In this method the Ti atoms, N atoms, O atoms and mized conformers, with and without alkoxy bridges,
C atoms (those involved in C–C coupling and the resulting from the bimolecular interaction of 14 are
methine carbon on isopropoxide) are considered as the summarized in table 1. Based on the very low energy
high layer and treated quantum mechanically. A mole- difference between 16-(± ) gauche and 16-meso anti
cular mechanics calculation is performed on the rest one might conclude that both (± ) and the meso iso-
of the system to understand the steric effects of the mers would be formed in nearly equal amounts.1 It
3d
isopropoxide which is defined as the low layer.
is interesting to note that the bimolecular intermedi-
Calculations were carried out to compare the rela- ate that involves alkoxy bridges is destabilized rela-
tive stabilities of alkoxy bridged trimeric biradical tive to the dinuclear species with out a bridge. The
species 9B with its non-bridged form 9. These stud- bridged meso isomer 16B-meso gauche is destabilized
ies reveal that bridged form 9B is more stable than the by about 17 kcal/mol compared to the most favourable
non-bridged form 9 by about 30 kcal/mol. During the non-bridged meso isomer 16-meso anti (entries 7 and
III
reductive coupling, if the imine 10 reacts with a Ti
species 8, then it would lead to the formation of a Ti
8, table 1). Similarly, the formation of the bridged
(± ) isomer 16B-(± )gauche is unfavourable by about
IV
species 14 (scheme 9). The trimeric biradical species 28 kcal/mol relative to its most stable non-bridged form
9
/9B could react with imines to give the trimeric species 16-(± ) gauche (entries 2, 3 and 9, table 1). Since exper-
1
5/15B (scheme 10). All intermediates (15/15B and 14) imental results show the exclusive formation of (± ) iso-
could lead to the coupled product 13.
mer, it is unlikely that intermediate 16 is involved in the
Two molecules of the Ti intermediate 14 are com- reaction.
IV
13d
IV
bined to give the C–C coupled Ti intermediates, 16-
If one considers a species such as 15, which does
(
± ) or 16-meso (scheme 11). The alkoxy bridged forms not contain an alkoxy bridge, the coupling would lead
Table 1. Relative energies of optimized conformers of 16 with and without alkoxy bridges.
Entry
Initial conformation
Optimized conformation
Energy (kcal/mol)
◦
◦
1
2
3
4
5
6
7
8
9
16-(± ) eclipsed (Dihedral angle = 0 )
16-(± ) eclipsed (Dihedral angle = 2 )
17.7
0.0
0.0
5.1
10.0
6.8
2.1
19.0
27.9
◦
◦
16-(± ) gauche (Dihedral angle = 60 )
16-(± ) gauche (Dihedral angle = 70 )
ꢁ
◦
◦
16-(± ) gauche (Dihedral angle = 120 )
16-(± ) gauche (Dihedral angle = 70 )
◦
◦
16-(± ) anti (Dihedral angle = 180 )
16-(± ) anti (Dihedral angle = 180 )
◦
◦
16- meso eclipsed (Dihedral angle = 0 )
16- meso gauche (Dihedral angle = 60 )
ꢁ
◦
ꢁ
◦
16- meso gauche (Dihedral angle = 120 )
16- meso gauche (Dihedral angle = 90 )
◦
◦
16- meso anti (Dihedral angle = 180 )
16- meso anti (Dihedral angle = 180 )
◦
◦
16B- meso eclipsed (Dihedral angle = 0 )
16B- meso gauche (Dihedral angle = 76 )
◦
◦
16B-(± )eclipsed (Dihedral angle = 0 )
16B-(± )gauche (Dihedral angle = 68 )

1350
Akshai Kumar and Ashoka G Samuelson
Table 2. Relative energies of optimized conformers 17 with and without alkoxy bridges.
Entry
Initial conformation
Optimized conformation
Energy (kcal/mol)
◦
◦
1
2
3
4
5
17-(± ) eclipsed (Dihedral angle = 0 )
17-(± ) gauche (Dihedral angle = 49 )
13.4
17.0
0.0
29.9
41
◦
◦
17- meso eclipsed (Dihedral angle = 0 )
17- meso gauche (Dihedral angle = 51 )
◦
◦
17B-(± ) eclipsed (Dihedral angle = 0 )
17B-(± ) eclipsed (Dihedral angle = 17 )
◦
◦
17B-(± ) gauche (Dihedral angle = 60 )
17B-(± ) gauche (Dihedral angle = 27 )
◦
◦
17B- meso eclipsed (Dihedral angle = 0 )
17B- meso gauche (Dihedral angle = 50 )
to the multinuclear intermediates 17-(± ) gauche or 17- based on Becke’s three-parameter functional.14 The
meso gauche. On the contrary, intermediate 15B which LANL2DZ basis set uses the effective core potentials
15
contains two alkoxy bridges, gives 17B-(± ) eclipsed (ECP) of Hay and Wadt. The nature of the stationary
or 17B-meso gauche (scheme 12). Due to the cyclic points is characterized by vibrational frequency calcu-
form of the intermediate (15/15B), only eclipsed and lations. The Gaussian 03 program package is used for
16
gauche conformations are possible. We now compare all calculations.
the stability of various conformations of intermediate
17 with and without alkoxy bridges (table 2). Unlike
the dinuclear intermediates discussed above, the forma- 3.2 Comparison of stability of bridged and
tion of alkoxy bridged (± ) isomer 17B-(± ) eclipsed non-bridged intermediates generated by the reaction of
is favoured by about 13 kcal/mol compared to its non- LVT titanium alkoxides and imines
bridged isomer 17-(± ) gauche (entries 1 and 3, table
The imine 10, titanium intermediates 8, 9, 16 and 17 are
modelled. All structures are optimized using the mul-
tilayer ONIOM model was used where the Ti atoms,
N atoms, O atoms and C atoms (involved in C–C cou-
pling and methine carbon on isopropoxide) are consi-
dered as the high layer and the rest of the atoms are
defined as a low layer. The high layer was optimized
using DFT at the B3LYP/LANL2DZ level of theory and
the low layer was optimized using molecular mechanics
calculation with the UFF forcefield. The Gaussian 09
program package was used for all multilayer ONIOM
2
). Surprisingly in the case of the meso isomer, the
non-bridged form, 17-meso gauche is stabilized by
5 kcal/mol relative to the alkoxy bridged isomer 17B-
2
meso gauche (entries 2 and 5, table 2). Thus, the alkoxy
bridged forms are favoured in the intermediates leading
to the formation of the (± ) coupled product, whereas
the favoured meso coupling involves non-bridged inter-
mediates. Thus, the most stable conformer 17B-(± )
eclipsed (entry 3, table 2) is the one that leads to (± )
diastereoselectivity. This is completely in accordance
with experimental results.
Thus, hybrid computational methods are useful in
determining the nature of intermediates in the reactions
of large molecules. In this case, it would not have been
possible to distinguish the two paths without recourse
to computational methods. Not only do they allow us to
distinguish between the dimeric and trimeric interme-
diates, they also throw light on the bridging nature of
alkoxides in these species.
17
calculations.
4
. Conclusion
Computational studies carried out at the B3LYP/
LANL2DZ level on model metal alkoxides are help-
ful in understanding the reactivity pattern and the reac-
tion path. There are distinct differences in the reactivity
of PhNCO and PhNCS towards metathesis. While
PhNCO undergoes mono insertion and double insertion
either in a head-to-tail or head-to-head fashion, PhNCS
exhibits only mono insertion and no double insertion.
Calculations attribute this difference to the energetics
of the reaction. Double insertion of PhNCO into tita-
nium(IV) alkoxides is highly favourable. On the other
3
. Computational details
3.1 Comparison of insertion and metathesis
reactivity of isocyanates and isothiocyanates catalysed
by titanium (IV) alkoxides
For theoretical studies, Ti(OMe) and Zr(OMe) are hand, this reaction is calculated to be an energetically
4
4
n
taken as model complexes in place of Ti(O Bu) and costly process for PhNCS. This renders PhNCS inac-
Zr(O Bu) where bulky Bu groups are replaced by Me tive towards double insertion and subsequent metathe-
4
n
n
4
groups for computational efficiency. All structures are sis reactions. Due to similar reasons, the metathesis of
optimized using the DFT method (B3LYP/LANL2DZ), carbon disulphide and carbodiimides fails to occur.

Computational tools.....reactions mediated by titanium(IV) isopropoxide
1351
Multilayer ONIOM calculations are used to investi-
gate the relative stabilities of multinuclear Ti intermedi-
ates that are involved in the diastereoselective coupling
of imines. Although, earlier studies show that reduc-
tion of imines by a trimeric biradical species is respon-
sible for the selectivity, it is not clear if the intermedi-
ates involve a bridged species. Careful examination of
the energies of the intermediates involved in the reac-
tion reveal the importance of alkoxy bridged multinu-
clear species in achieving diastereoselective coupling.
It is important to note that one of the precursors based
on the trimeric biradical is stabilized to a great extent
by the presence of alkoxy bridges. Calculations show
that the non-bridged forms are favoured in the dinu-
clear intermediates! Again, this is a counter intuitive
result as one would think alkoxy bridges are impor-
tant in all multinuclear Ti species. Moreover, this binu-
clear interaction is likely to result in a mixture of both
diastereomers. Alkoxy bridging is only important in the
trimeric species where (± ) coupling is favoured. Given
the fact that there is an exclusive formation of (± ) iso-
mers experimentally, these calculations strongly indi-
cate the involvement of intermediates like 17/17B in
the reaction. Once again, we find that computational
tools are useful in determining the reaction pathway
involving titanium-isopropoxide complexes.
Chem. 100 19357; (g) Dapprich S, Komaromi I, Byun
K S, Morokuma K, Frisch M J 1999 J. Mol. Struct.:
THEOCHEM 461–462 1
. (a) Davidson E R 2000 Chem. Rev. 100 351; (b) Bruice
T C 2006 Chem. Rev. 106 3119; (c) Balcells D, Clot E
and Eisenstein O 2010 Chem. Rev. 110 749; (d) Bifulco
G, Dambruoso P, Gomez-Paloma L and Riccio R 2007
Chem. Rev. 107 3744; (e) Proft F D and Geerlings P 2001
Chem. Rev. 101 1451; (f) Kraka E and Cremer D 2010
Acc. Chem. Res. 43 591; (g) Zhao Y and Truhlar D G
2
2008 Acc. Chem. Res. 41 157
3
. (a) Schwabe T and Grimme S 2008 Acc. Chem. Res.
41 569; (b) Clementi E, Corongiu G, Bahattacharya D,
Feuston B, Frye D, Preiskorn A, Rizzo A and Xue W
1
991 Chem. Rev. 91 679; (c) Karadakov P B 2001 Annu.
Rep. Prog. Chem. Sect. C 97 61
4
. (a) Morokuma K 2007 Bull. Chem. Soc. Jpn. 80 2247;
(
b) Piyaauksornsak S, Tangthongkul T, Wanbayor R,
Wanno B and Ruangpornvisuti V 2009 Struct. Chem. 20
767; (c) Ananikov V P, Musaev D G and Morokuma K
2010 J. Mol. Catal. A: Chem. 324 104
5
. (a) Bo C and Maseras F 2008 Dalton Trans. 2911; (b)
Basiuk V A and Basiuk E V 2004 Nanotechnology 2 395
. (a) Svelle S, Visur M, Olsbye U, Saepurahman and
Bjørgen M 2011 Top Catal. 54 897; (b) North M, Usanov
D L and Young C 2008 Chem. Rev. 108 5146; (c)
Bakalova S M and Kaneti J 2008 J. Phys. Chem. A
6
112 13006; (d) Dudding T and Houk K N 2004 Proc.
Natl. Acad. Sci. USA 101 5770; (e) Cantillo D, Avalos
M, Babiano R, Cintas P, Jimenez J L and Palacios J
C 2011 Org. Biomol. Chem. 9 7638; (f) Ishikawa A,
Nakao Y, Sato H and Sakaki S 2010 Dalton Trans. 3279;
(
g) Cowley R E, Eckert N A, Vaddadi S, Figg T M,
Cundari T R and Holland P L 2011 J. Am. Chem. Soc.
33 9796; (h) Wang C-M, Wang Y-D, Liu H-X, Xie Z-K
Acknowledgements
1
AGS thanks the Department of Science and Technol-
ogy (DST), New Delhi for the award of a research
grant and AK gratefully acknowledges for Post Doc-
toral Fellowship from Indian Institute of Science (IISc),
Bangalore. Authors also thank Prof. ED Jemmis, Dr.
S De and Dr. S Dinda and M Thenraj for their help
and suggestions. Thanks are also due to Supercom-
puter Education and Research Centre (SERC) of IISc
for computational facilities.
and Liu Z-P 2010 J. Catal. 271 386
7
. (a) Pavan Kumar P N V and Jemmis E D 1988 J. Am.
Chem. Soc. 110 125; (b) Jemmis E D and Giju K T 1997
Angew. Chem. Int. Ed. Engl. 36 606
8. (a) Jemmis E D and Giju K T 1998 J. Am. Chem. Soc.
1
2
20 6952; (b) Jemmis E D, Phukan A K and Giju K T
002 Organometallics 21 2254
9
. Kaleta K, Ruhmann M, Theilmann O, Beweries T, Roy
S, Arndt P, Villinger A, Jemmis E D, Schulz A and
Rosenthal U 2011 J. Am. Chem. Soc. 133 5463
1
0. (a) Schofield A D, Nova A, Selby J D, Schwarz A D,
Clot E and Mountford P 2011 Chem. Eur. J. 17 265;
(b) Tiong P J, Nova A, Groom L R, Schwarz A D,
Selby J D, Schofield A D, Clot E and Mountford P
Supplementary information
2
011 Organometallics 30 1182–1201; (c) Schofield A D,
The supplementary information can be seen in
www.ias.ac.in/chemsci.
Nova A, Selby J D, Manley C D, Schwarz A D, Clot E
and Mountford P 2010 J. Am. Chem. Soc. 132 10484;
(
d) Tiong P-J, Schofield A D, Selby J D, Nova A, Clot E
and Mountford P 2010 Chem. Commun. 46 85; (e) Selby
J D, Schulten C, Schwarz A D, Stasch A, Clot E, Jones
C and Mountford P 2008 Chem. Commun. 5101
References
1
. (a) Carter E A and Rossky E J 2006 Acc. Chem. Res. 39 11. (a) Guiducci A E, Boyd C L, Clot E and Mountford
7
1; (b) Lin Z 2010 Acc. Chem. Res. 43 602; (c) Schröder
D, Shaik S and Schwarz H 2000 Acc. Chem. Res. 33
39; (d) Scuseria G E 1999 J. Phys. Chem. A 103 4782;
P 2009 Dalton Trans. 5960; (b) Bolton P D, Feliz M,
Cowley A R, Clot E and Mountford P 2008
Organometallics 27 6096
1
(
e) Maseras F and Morokuma K 1995 J. Comput. Chem. 12. (a) Ghosh R, Nethaji M and Samuelson A G 2005
1
6 1170; (f) Svensson M, Humbel S, Froese R D J,
J. Organomet. Chem. 690 1282; (b) Ghosh R and
Samuelson A G 2005 Chem. Commun. 115 2017; (c)
Matsubara T, Sieber S and Morokuma K 1996 J. Phys.

1352
Akshai Kumar and Ashoka G Samuelson
Ghosh R, Nethaji M and Samuelson A G 2003 Chem.
Commun. 113 2556; (d) Kumar A, De S, Samuelson
A G and Jemmis E D 2008 Organometallics 27 955;
Cammi R, Pomelli C, Ochterski J W, Ayala P Y,
Morokuma K, Voth G A, Salvador P, Dannenberg J J,
Zakrzewski V G, Dapprich S, Daniels A D, Strain M C,
Farkas O, Malick D K, Rabuck A D, Raghavachari
K, Foresman J B, Ortiz J V, Cui Q, Baboul A G,
Clifford S, Cioslowski J, Stefanov B B, Liu G,
Liashenko A, Piskorz P, Komaromi I, Martin R L, Fox
D J, Keith T, Al Laham M A, Peng C Y, Nanayakkara A,
Challacombe M, Gill P M W, Johnson B, Chen W, Wong
M W, Gonzalez C and Pople J A 2004 Wallingford CT:
Gaussian, Inc
(
e) Kumar A and Samuelson A G 2010 J. Organomet.
Chem. 695 338; (f) Kumar A and Samuelson A G 2011
J. Chem. Sci. 123 29; (g) Ghosh R 2004 Ph. D Thesis
Indian Institute of Science; (h) Kumar A 2009 Ph. D
Thesis, Indian Institute of Science
1
3. (a) Qin S, Hu C, Yang H and Su Z 2008 J. Org.
Chem. 73 4840; (b) Martínez S, Ramos J, Cruz V L and
Martínez-Salazar J 2006 Polymer 47 883; (c) Kumar A
and Samuelson A G 2010 Chem. Asian J. 5 1830; (d) 17. Frisch M J, Trucks G W, Schlegel H B, Scuseria G E,
Kumar A and Samuelson A G 2011 Eur. J. Org. Chem.
51
4. (a) Becke A D 1993 J. Chem. Phys. 98 5648; (b) Becke
A D 1988 Phys. Rev. A 38 3098; (c) Lee C, Yang W and
Parr R G 1988 Phys. Rev. B 37 785
Robb M A, Cheeseman J R, Scalmani G, Barone V,
Mennucci B, Petersson G A, Nakatsuji H, Caricato M,
Li X, Hratchian H P, Izmaylov A F, Bloino J Zheng G,
Sonnenberg J L, Hada M, Ehara M, Toyota K, Fukuda
R, Hasegawa J, Ishida M, Nakajima T, Honda Y, Kitao
O, Nakai H, Vreven T, Montgomery J A, Peralta Jr,
J E, Ogliaro F, Bearpark M, Heyd J J, Brothers E,
Kudin K N, Staroverov V N, Kobayashi R, Normand J,
Raghavachari K, Rendell A, Burant J C, Iyengar S S,
Tomasi J, Cossi M, Rega N, Millam J M, Klene M,
Knox J E, Cross J B, Bakken V, Adamo C, Jaramillo J,
Gomperts R, Stratmann R E, Yazyev O, Austin A J,
Cammi R, Pomelli C, Ochterski J W, Martin R L,
Morokuma K, Zakrzewski V G, Voth G A, Salvador P,
Dannenberg J J, Dapprich S, Daniels A D, Farkas O,
Foresman J B, Ortiz J V, Cioslowski J and Fox D J 2009
Gaussian 09, Revision A.02, Wallingford CT: Gaussian,
Inc
9
1
1
1
5. (a) Hay P J and Wadt W R 1985 J. Chem. Phys. 82 270;
(
(
b) Wadt W R and Hay P J 1985 J. Chem. Phys. 82 284;
c) Hay P J and Wadt W R 1985 J. Chem. Phys. 82 299
6. Gaussian 03, Revision C.02, Frisch M J, Trucks G W,
Schlegel H B, Scuseria G E, Robb M A, Cheeseman
J R, Montgomery Jr. J A, Vreven T, Kudin K N, Burant
J C, Millam J M, Iyengar S S, Tomasi J, Barone V,
Mennucci B, Cossi M, Scalmani G, Rega N, Petersson
G A, Nakatsuji H, Hada M, Ehara M, Toyota K, Fukuda
R, Hasegawa J, Ishida M, Nakajima T, Honda Y, Kitao
O, Nakai H, Klene M, Li X, Knox J E, Hratchian
H P, Cross J B, Bakken V, Adamo C, Jaramillo J,
Gomperts R, Stratmann R E, Yazyev O, Austin A J,